Helping People grow!

Editor’s Note.
The impact of child care on children’s development is both complex and controversial. Research findings since the 1980s demonstrate the important impact of child care experiences on children’s development. Indeed, consensus has now emerged among investigators that child care quality is linked to the well-being, skills, and subsequent adjustment of young children. Yet, no such consensus has been reached on the influence of quantity of exposure to child care during the early years.
Findings from the National Institute of Child Health and Human Development (NICHD) Study of Early Child Care and Youth Development, the most comprehensive, longitudinal study in the United States of the relationships between child care experiences and children’s development, will inform the discussion and debate on child care and children’s development. As reviewed by Belsky in his Commentary, the NICHD study reports that children with greater exposure to nonmaternal care during the first 4 1/2 years of life were rated by their preschool and kindergarten teachers as more aggressive, assertive, and defiant than same-age children with less exposure. Researchers are now speculating as to the implications of such findings.
Belsky, a member of the NICHD Early Child Care Research Network, argues that quantity does indeed count. Other researchers (and Network members) are more cautious in their interpretation, citing the lack of clinical significance of ratings of problematic behaviors, the partially mediating influence of quality of care on quantity effects, and the more positive effects of exposure on peer interactions.
The Journal is pleased to advance the discourse (and debate) on this important topic by publishing this solicited Commentary. We have invited a response to this Commentary for subsequent publication from another expert in this field. We encourage readers to review the publications now emerging from the NICHD Early Child Care Research Network. As always, we welcome readers’ thoughts and responses. Paul H. Dworkin, M.D. Editor

In the mid- to late 1980s a major controversy erupted in both the scientific literature and the popular press when Belsky’s1–3 analysis of child care research produced the conclusion that early and extensive nonmaternal care carried risks in terms of increasing the probability of insecure infant-parent attachment relationships and promoting aggression and noncompliance during the toddler, preschool, and early primary school years. Widespread critiques of Belsky’s analysis called attention to much of the research cited by Belsky and its failure to take into consideration child-care quality and control for background factors likely to make children with varying child-care experiences developmentally different in the first place. 4–8
In the 15 years since Belsky1 first advanced his controversial conclusions about the effects of child care—after his previous reviews of the literature revealed little evidence that early child care carried developmental risks9 but that quality of care is a most important determinant of children’s development10—a great deal of research on child care and maternal employment in the opening years of the child’s life has been reported. It is noteworthy that what Belsky1 characterized as ‘‘a slow steady trickle of disconcerting evidence’’ pertaining to the effects of
nonmaternal care initiated in the first year of life on socioemotional development, especially on a full-time or near full-time basis,2,3 has continued unabated over the past decade and a half, perhaps even increasing in flow.11

Consider in this regard the following research findings:

(1) Work by Clark, Hyde, Essex and Klein12 on almost 200 mothers showing that earlier return to paid employment following the child’s birth predicts more negative maternal affect and behavior directed toward the child and higher levels of infant dysregulation and irritability;

(2) work by Braungart-Rieker, Courtney, and Garwood13 replicating earlier findings14,15 showing sons to be at elevated risk of developing insecure attachments to their fathers when
mothers are employed on a full- or near full-time basis in the child’s first year of life;

(3) work by Vandell and Corasaniti16 on more than 200 white 8 to 9 year olds showing that children whose full time (>30 hours/week) nonmaternal care is initiated in the first year of life and continues at this high level until school entry are rated by mothers and teachers as more noncompliant and as getting along with peers more poorly than other children and by
peers as being less likable;

(4) work by Bates and associates17 on almost 600 5 to 6 year olds entering kindergarten showing that children who spend more time in child care during their first 5 years of life score lower on a composite measure of positive adjustment (i.e., peer popularity, teacher-rated peer competence) and higher on a composite measure of negative adjustment (i.e., teacher rated behavior problems, peer dislike, observed aggression) than children with less child care experience;

(5) work by Hofferth18 on 519, 3 to 4 year olds participating in the nationally representative Panel Study of Income Dynamics showing initiation of child care during the first year of life to be associated with significantly higher scores on an index of behavior problems, especially aggressive problems; and, finally,

(6) work by Han, Waldfogel, and Brooks-Gunn19 using data on 138 white children whose parents were enrolled in the National Longitudinal Study of Youth showing that resumption of maternal employment in the first 9 months of the child’s life predicts higher levels of motherreported externalizing problems at age 7 to 8 years. Besides providing evidence consistent with Belsky’s1 much-contested conclusion that developmental risks are
associated with nonmaternal child care initiated in the first year, especially on a full- or near full-time basis,2,3 and his subsequent observation that it is such full- or near full-time
care that begins in the first year and continues unabated until school entry that is principally responsible for the risks detected (i.e., early, extensive, and continuous care),11,20 the
investigations just cited are important for two reasons.

First, in all cases, the cited studies addressed one of the core limitations of previous work in that a variety of background factors were controlled before estimating effects of early child care and/or maternal employment. This analytic strategy reduces the prospect that putative effects of child care reflect nothing more than pre-existing differences among families that rely upon varying amounts of child care.
The fact that so much research that has emerged since the inception of what Karen21 called ‘‘the child care wars’’ has overcome one of the major weaknesses of the body of work
on which Belsky1 based his original conclusions, yet still yields results consistent with his observations, does not mitigate the fact that most of this more recent work highlighting risks associated with early child care is still limited by its failure to take into consideration the extent to which caregivers are attentive, responsive, stimulating, and affectionate. This lacuna is of critical importance because the mantra of most developmentalists and child care advocates
in response to Belsky’s1–3 observations has been that ‘‘it is not the quantity (or timing) of child care/maternal employment that is developmentally significant, but the quality of care.’’5–7 That is, the only reason that lots of time spent in nonmaternal child care beginning in the first year is at all associated with lower quality parent-child interaction, increased rates of insecure attachment, and/or higher levels of aggression and disobedience is that children are experiencing low-quality child care that has gone unmeasured in all too much research.
Fortunately, the National Institute of Child Health and Human Development (NICHD) Study of Early Child Care, a longitudinal investigation of more than 1000 children born in 10 different United States communities and followed from birth into their middle-childhood years, was designed to address this very issue (and many others).

And the results to date are most revealing. Even after controlling for a host of background factors distinguishing families and children, and even after taking into consideration the observed quality of child care that children experienced at 6, 15, 24, 36, and 54 months of age, as well as the type of care experienced (e.g., nanny, family day care home, center), the NICHD Study finds that quantity or dosage of child care predicts a host of developmental outcomes, and, typically, in a manner consistent with Belsky’s1 original risk-factor conclusion and/or subsequent observation that it is a cumulative history of lots of time in child care that carries developmental risks.11,20

Consider first results pertaining to security of attachment to mother: When children averaged more than just 10 hours per week in care and their mothers provided relatively insensitive care to them, children were at increased risk of developing insecure attachments, as measured in the Strange Situation at 15 months and at 36 months of age.22 Consider next observed patterns of mother-child interaction.
When children spent more time in nonmaternal care, mothers were less sensitive in interacting with their child at 6, 15, 24, and 36 months of age and children were less positively engaged with and responsive to their mothers.23 In fact, close examination of the data suggests that quantity of child care may have first affected mothering and, only thereafter and perhaps thereby, child behavior. This is because more time in nonmaternal care first predicted less
sensitive mothering when infants were 6 months old and then more negative mothering by 15 months, before predicting less positive engagement of mother by the child at ages 24 and 36 months.24

Finally, consider assessments of socioemotional adjustment. When children spent more time in child care through their first 2 years of life, they were less cooperative, according to maternal reports, and evinced more problem behavior, according to caregiver reports, at 24 months of age. No such significant findings emerged, however, when amount of care through the first 3 years was used to predict child behavior at age 3 years. But, by the time children were 4.5 years of age, more time in nonmaternal care through the first 54 months of life again forecast higher levels of caregiver-reported (but not mother-reported) problem behavior.25 And, by the time children were in kindergarten, approximately a year later, more time in care predicted
higher levels of externalizing problems as reported by mothers and teachers alike.26 In other words, the relation between more time in child care and more problem behavior emerged when ratings by three separate sets of raters, each of whom observed children in a different setting, were examined: caregivers in child care, mothers at home, and teachers in school.

Just as important as this replicated relation between quantity of child care and problem behavior was the fact that across all raters, more time in care predicted high externalizing scores, that is, scores one or more standard deviations above the mean.26 At age 4.5 years, whereas 5% of children who averaged under 10 hours per week of nonmaternal care across their first 54 months of life received such elevated externalizing problem scores on the basis of caregiver ratings, the rate of such high levels of problems was more than three times as great (16%) among agemates who averaged 30 or more hours per week of care across their first 4.5 years of life. At kindergarten age, the corresponding rates of high levels of problems were 9% and 17% for the two extreme groups that had experienced limited and full-time care throughout their infant, toddler, and preschool years.

Two other features of the NICHD results are notable. First, no quantity threshold could be detected at which more versus less care had a noticeably greater or lesser impact on
problem behavior.26 Thus, the relation between amount of nonmaternal care and externalizing problems reflected a constant dose-response relationship: as quantity of care increased, so did problem behavior. Importantly, item-level analysis revealed that it was not the case, as Clarke-Stewart27 speculated might be so, that children with extensive child-care histories were simply more independent and assertive than other children. Rather, in the NICHD research, they were found to show evidence of neediness (e.g., demands a lot of attention, demands must be met immediately, easily jealous), assertiveness (e.g., bragging/boasting, argues a lot), disobedience/defiance (e.g., talks out of turn, disobedient at school, defiant-talks back to staff,
disrupts school discipline), and aggression (e.g., gets into many fights, cruelty-bullying-meanness, physically attacks others, destroys own things).

Clearly, these results from the NICHD Study are rather consistent with much of the work cited earlier highlighting risks associated with lots of time spent in early child care. In other words, the NICHD Study is by no means the first, not even since the outbreak of the so-called ‘‘day care wars,’’ to report evidence linking early entry into and/or lots of time in child care with insecure attachment, less harmonious patterns of mother-child interaction, or higher levels of
problem behavior, especially aggression. What is new and noteworthy, though, is that the NICHD results reveal that such findings cannot be explained by quality (or type) of child care. Also noteworthy is that all the NICHD results summarized herein held true across the sample; factors like family economic status, marital status, and maternal education did not moderate the effects of quantity of child care detected in this research.

Especially important to understanding the effect of early child care on problem behavior in the NICHD Study are the effect-size comparisons carried out to illuminate the effect of quantity of care relative to that of two well-known correlates of children’s development: economic impoverishment and parenting skill. As it turned out, the effect size of cumulative hours in nonmaternal care across the first 4.5 years of life was virtually the same magnitude as that
of poverty and substantially larger than that of sensitiveresponsive parenting.25 To be noted, however, is that they were modest in magnitude when viewed in absolute, rather than relative, terms. The contrast between children averaging less than 10 hours per week and those averaging 30 or more hours per week on externalizing problems at 54 months yielded an effect size, d = .38, for example.

Conclusion: On the basis of the evidence summarized, it would seem no longer tenable for developmental scholars and child-care advocates to deride the notion that early, extensive, and
continuous nonmaternal care of the kind available in most communities poses some developmental risks for young children and perhaps the larger society as well. Nor would
it seem any longer tenable to argue that disconcerting evidence highlighting risks associated with high dosages of child care are simply a function of low-quality care. As indicated in the title of this commentary, quantity counts when it comes to understanding the effects of early child
care on socioemotional development. So, too, of course, does quality of care, because much uncited evidence from the NICHD Study and many other investigations indicates that children do better when they experience high- rather than low-quality child care, especially in terms of their cognitive and language development. But these effects of quality of care are, overwhelmingly, independent of those of quantity of care.

The potential risks associated with both lots of time in child care and with low-quality child care need to be contextualized. First, it should be appreciated that the seemingly negative effects of high dosages of child care, just like the positive effects of good quality of care, are by
no means large; in fact, they are rather modest. But these potentially small effects must be considered in the context of a society in which more and more children are spending
more and more time at younger and younger ages in nonmaternal care and in which all too much of that child care is of distinctly limited quality. This means that even small effects, when experienced by many children, may have broad-scale consequences. After all, many of the most important risk behaviors from a public health perspective have low or moderate relative risk but are multiplied in importance because of their wide prevalence and links to problematic outcomes.28 This may be the case for early, extensive, and continuous nonmaternal care and for low-quality care.

Ironically, this state of affairs leads me to draw virtually the same policy-related conclusions that I drew a decade ago,3 after first choosing not to draw any to keep separate scientific analysis and policy inference because I felt strongly that the latter is and should be shaped as much by
social values as by empirical evidence.1,2 First, it seems to me that the data considered in this essay should encourage the expansion of parental leaves, preferably paid and ideally as lengthy as they are in some Scandinavian countries, or other strategies for reducing the time children spend in nonmaternal care across the infant, toddler, and preschool years (e.g., part-time employment). Such a change would be consistent with the stated preferences of American parents of young children who, national survey data reveals, believe that having a full-time parental presence in the home is what is best for very young children and that the care and attention children get from even ‘‘a top-notch day care center’’ is just not as good as what they would get at home with a parent.29

Relatedly, tax policies should support families rearing infants and young children in ways that afford parents the freedom to make childrearing arrangements that they deem to be best for their child, thereby reducing the economic coercion that necessitates many to leave the care of their children to others when they would rather not. Finally, given the clear benefits of high-quality child care, its expansion seems called for as well. Of significance is that all
of these conclusions could be justified on humanitarian grounds alone.

Jay Belsky Institute for the Study of Children, Families, and Social Issues Birkbeck College, University of London London, United Kingdom

REFERENCES
1. Belsky J: Infant day care: A cause for concern? Zero to Three 6:1–9, 1986
2. Belsky J: The effects of infant day care reconsidered. Early Child Res Q 3:235–272, 1988
3. Belsky J: Developmental risks associated with infant day care: Attachment insecurity, noncompliance, and aggression? In Chehrazi S (ed): Psychosocial Issues in Day Care. Washington, DC, American Psychiatric Press, 1990, pp 37–68
4. Fox N, Fein G (eds): Infant day care: The current debate. Norwood, NJ, Ablex, 1990
5. McGurk H, Caplan M, Hennessy E, Moss P: Controversy, theory, and
social context in contemporary day care research. J Child Psychol
Psychiatry 34:3– 23, 1993
6. Phillips D, McCartney K, Scarr S, Howes C: Selective review of
infant day care research: A cause for concern. Zero to Three 7:18–21,
1987
7. Scarr S, Phillips D, McCartney K: Facts, fantasies, and the future of
child care in the United States. Psychol Sci 1:26–35, 1989
8. Thompson R: The effects of infant day care through the prism of
attachment theory. Early Child Res Q 3:273– 282, 1988
9. Belsky J, Steinberg L: The effects of day care: A critical review.
Child Dev 49:929–949, 1978
10. Belsky J: Two waves of day care research: Developmental effects and
conditions of quality, in Ainslie R (ed): The Child and the Day Care
Setting. New York, NY, Prager, 1984, pp 1– 34
11. Belsky J: Emanuel Miller Lecture: Developmental risks (still) associated
with early child care. J Child Psychol Psychiatry, 42:845– 859,
2001
12. Clark R, Hyde J, Essex M, Klein M: Length of maternity leave and
quality of mother-infant interactions. Child Dev 68:364–383, 1997
13. Braungart-Rieker JM, Courtney S, Garwood M: Mother- and fatherinfant
attachment: Families in context. J Fam Psychol 13:535– 553,
1999
14. Belsky J, Rovine M: Nonmaternal care in the first year of life
and security of infant-parent attachment. Child Dev 59:157– 167,
1988
15. Chase-Lansdale P, Owen M: Maternal employment in a family
context. Child Dev 58:1505– 1512, 1987
16. Vandell D, Corasaniti H: Child care and the family: Complex
contributors to child development. New Dir Child Dev 49:23– 37,
1990
17. Bates J, Marvinney D, Kelly T, Dodge K, Bennett R, Pettit G: Childcare
history and kindergarten adjustment. Dev Psychol 30:690– 700,
1994
18. Hofferth S: Child care in the first three years of life and preschoolers’
language and behavior. Paper presented at the Biennial Meeting of
the Society for Research in Child Development, Albuquerque, NM,
April 1999
19. Han W, Waldfogel J, Brooks-Gunn J: The effects of early maternal
employment on later cognitive and behavioral outcomes. J Marriage
Fam 63:336–354, 2001
20. Belsky J: The effects of infant day care: 1986–1994. Invited plenary
address to the British Psychological Association Division of
Developmental Psychology, University of Portsmouth, England,
Sept 4, 1994
21. Karen R: Becoming Attached: Unfolding the Mystery of the Infant-
Mother Bond and Its Impact on Later Life. New York, NY, Warner
Books, 1994
22. NICHD Early Child Care Research Network: The effects of infant
child care on infant-mother attachment security. Child Dev
68:860– 879, 1997
23. NICHD Early Child Care Research Network: Child care and
mother-child interaction in the first 3 years of life. Dev Psychol
35:1399– 1413, 1999
24. NICHD Early Child Care Research Network: Mother-child interaction
and cognitive outcomes associated with early child care:
Results of the NICHD study. Poster symposium presented at the
Biennial Meeting of the Society for Research in Child Development,
Washington, DC, April 14, 1997
25. NICHD Early Child Care Research Network: Early child care and
children’s development prior to school entry. Presented at the
Biennial Meeting of the Society for Research in Child Development,
Minneapolis, MN, April 19, 2001
26. NICHD Early Child Care Research Network: Further explorations of
the detected effects of quantity of early child care on socioemotional
adjustment. Paper presented at the Biennial Meeting of the Society
for Research in Child Development, Minneapolis, MN, April 19,
2001
27. Clarke-Stewart K: Infant day care: Maligned or malignant? Am
Pscyhol 44:266– 273, 1989
28. Jeffrey R: Risk behaviors and health: Contrasting individual and
population perspectives. Am Psychol 44:1194–1202, 1989
29. Farkas S, Duffett A, Johnson J: Necessary compromises: How
parents, employers and children’s advocates view child care today.
New York, NY, Public Agenda, 2000
See related Journal Article Review, p 186, ‘‘NICHD Early Child Care Research Network:
Child-care and family predictors of preschool attachment and stability from infancy.’’
Dev Psychol 37:847, 2001.
170 BELSKY JDBP/June, Vol. 23, No. 3

Commentary
Address for reprints: Jay Belsky, Institute for the Study of Children, Families, and Social Issues, Birkbeck College, University of London, 7 Bedford Square, London WC1B 3RA, United Kingdom; e-mail: j.belsky@bbk.ac.uk; fax: 44 20 7436 2750. Work on this paper was supported by a grant from the National Institute of Child Health and Human Development.
167 0196-206X/00/2303-0167
Developmental and Behavioral Pediatrics Vol. 23, No. 3, June 2002 Copyright 2002 by Lippincott Williams & Wilkins, Inc. Printed in
U.S.A. #